US government mandates shutdown of Anthropic's Mythos and Fable 5 AI models over cyber security risks

The U.S. government forced Anthropic to decommission its Mythos and Fable 5 models on June 12, 2026, citing unacceptable cyber-offensive potential. Anthropic disputes the necessity of the shutdown, characterizing the vulnerability as a narrow, non-universal jailbreak. The incident highlights the absence of standardized regulatory frameworks for AI safety and the friction between private sector development and federal security mandates.